
\begin{table}
\caption{Regression Results for Vignette Analysis}
\begin{center}
\begin{scriptsize}
\begin{threeparttable}
\begin{tabular}{l D{.}{.}{4.6} D{.}{.}{4.6} D{.}{.}{4.6} D{.}{.}{4.6}}
\toprule
 & \multicolumn{2}{c}{Deservingness} & \multicolumn{2}{c}{Redistribution} \\
\cmidrule(lr){2-3} \cmidrule(lr){4-5}
 & \multicolumn{1}{c}{(1) Cont. } & \multicolumn{1}{c}{(2) Dich.} & \multicolumn{1}{c}{(3) Cont.} & \multicolumn{1}{c}{(4) Dich.} \\
\midrule
Nazi past               & -0.785^{***}            & -0.243^{***}            & 0.672^{***}             & 0.227^{***}             \\
                        & (0.024)                 & (0.010)                 & (0.024)                 & (0.010)                 \\
Apology                 & -0.520^{***}            & -0.173^{***}            & 0.494^{***}             & 0.165^{***}             \\
                        & (0.024)                 & (0.010)                 & (0.024)                 & (0.010)                 \\
\midrule
Individual FE           & \multicolumn{1}{c}{yes} & \multicolumn{1}{c}{yes} & \multicolumn{1}{c}{yes} & \multicolumn{1}{c}{yes} \\
Num. obs.               & 6006                    & 6006                    & 6006                    & 6006                    \\
R$^2$ (full model)      & 0.780                   & 0.744                   & 0.782                   & 0.740                   \\
R$^2$ (proj model)      & 0.217                   & 0.146                   & 0.173                   & 0.126                   \\
Adj. R$^2$ (full model) & 0.670                   & 0.615                   & 0.672                   & 0.610                   \\
Adj. R$^2$ (proj model) & -0.175                  & -0.281                  & -0.241                  & -0.312                  \\
Num. groups: caseid     & 2002                    & 2002                    & 2002                    & 2002                    \\
\bottomrule
\end{tabular}
\begin{tablenotes}[flushleft]
\tiny{\item\textit{Note:} Linear regression (OLS) with individual fixed-effects; unit of analysis are vignettes (three per individual respondent); reference category = baseline vignette. Models 1 and 3 use continuous outcome variable, ranging from 1 (strongly disagree) to 5 (strongly agree); models 2 and 4 dichotomized outcome variables with 1 indicating (strong) agreement and 0 all other response options. (*=.05, **=.01, ***=.001)}
\end{tablenotes}
\end{threeparttable}
\end{scriptsize}
\label{table:results1}
\end{center}
\end{table}
